Abstract
Background:
Women have lower pulse pressure (PP) than men as young adults, but higher PP after a midlife PP nadir. This PP reversal is often attributed to lower estrogen levels at menopause. However, a detailed analysis of hemodynamic changes around the time of menopause has not been performed.
Methods:
Relations between the timing of menopause and sex differences in PP were assessed in the Framingham Heart Study cohorts (N=6760 unique participants) across three visits spanning 14 years. Women were categorized into 4 groups by age at menopause: persistently premenopausal or postmenopausal with early (<25th percentile, age <48 years), average (25th to 74th percentile, age 48 to 53), or late (≥75th percentile, age >53 years) menopause age.
Results:
PP decreased with increasing age in women and men under 40 years of age, being lower in women than in men. PP increased with age in women and men after midlife and was higher in women than in men after age 60. Nadirs in midlife PP, adjusted for vascular risk factors, occurred at 47 years of age in men and at 39, 38, 37, and 37 years of age in the four menopause groups, respectively, with nadirs in the postmenopausal groups occurring an average of 6, 14, and 19 years before menopause.
Conclusions:
PP transitions from a falling to a rising profile approximately a decade earlier in women than in men, and up to 2 decades before the onset of menopause in women, suggesting that factors other than age of menopause are involved.
Keywords: Arterial stiffness, sex characteristics, menopause, blood pressure, longitudinal studies
Graphical Abstract
Introduction
Wide pulse pressure (PP), the difference between systolic and diastolic blood pressure, is a marker of vascular stiffness that is strongly associated with incident cardiovascular disease (CVD).1 Despite its prognostic significance, PP is neither routinely monitored nor specifically addressed in clinical practice guidelines. Current evidence suggests that the primary mechanism by which PP widens is impaired coupling between pulsatile blood flow and the intrinsic properties of the aorta, leading to a larger forward pressure wave amplitude (FWA).2 Characteristic impedance (Zc), which assesses the pulsatile pressure-flow relation in the proximal aorta, is closely related to FWA and PP.2,3 Carotid-femoral pulse wave velocity (CFPWV) is a measure of the average wall stiffness of the aorta. CFPWV increases progressively with age across the adult lifespan. In contrast, Zc, FWA, and PP fall with advancing age through midlife, in parallel with an increase in aortic diameter, and then increase dramatically thereafter, especially in women, in parallel with an accelerated increase in CFPWV and a blunted increase in aortic diameter in men and women.4,5 These findings suggest that aortic properties, including lumen diameter and wall stiffness, play a major role in the development of elevated PP.6
Sex-related differences in PP are well-documented across the lifespan. Women generally demonstrate lower PP than men during early adulthood; however, this relation reverses with advancing age.6,7 The accelerated increase in PP after midlife in women has been attributed to the decline in circulating estrogen, as prior experimental and clinical studies have suggested favorable effects of estrogen on aortic size and proximal aortic stiffness. However, longitudinal clinical data from large samples on the relations between vascular health and age at menopause are limited.8–13 To address this knowledge gap, we analyzed longitudinal data from the Framingham Heart Study Cohort to elucidate the association between age at menopause and age-related measures of arterial stiffness and excess pressure pulsatility, including Zc, CFPWV, FWA, and PP.
Methods
FHS materials and data are publicly available upon request at the National Heart, Lung, and Blood Institute’s Biologic Specimen and Data Repositories Information Coordinating Center https://biolincc.nhlbi.nih.gov/studies/fhs/.
Participants.
The sample for the present study was drawn from the Framingham Offspring (N=3204), New Offspring Spouse (N=103), and Third Generation (N=4095) cohorts, and the racially and ethnically diverse Omni-1 (N=356) and Omni-2 (N=410) cohorts. Participants who attended ≥1 of 3 sequential examinations (1, 2, or 3), including Offspring Exams 8-10, Omni-1 Exams 3-5, and Third Generation, New Offspring Spouse and Omni-2 Exams 1-3, were eligible for inclusion. Of the 8,168 who attended a hemodynamic evaluation at visits 1, 2, or 3, participants were excluded for missing or incomplete covariate and tonometry data (N=511) or for incomplete data on menopause status or menopause resulting from secondary causes (N=897). The final analysis included N=6760 unique participants (Supplemental Figure S1). Each participant could be represented in the dataset up to three times, corresponding to 14520 total visits across 14 years. Out of 6760 unique participants, 6211 attended the hemodynamic evaluation at visit 1. The remaining 549 patients attended at least one of visits 2 and 3. All protocols were approved by the Boston University Medical Center Institutional Review Board, and participants provided written informed consent.
Clinical Evaluation.
Medical history, physical examination, and electrocardiography were performed routinely at each examination. Prevalent CVD was defined as presence of coronary heart disease, cerebrovascular disease, heart failure, or intermittent claudication. Serum cholesterol and triglyceride levels were measured from fasting blood samples. Diabetes mellitus was defined as a fasting glucose level ≥126 mg/dL, a random glucose level >200 mg/dL, or the use of hypoglycemic medications. Body mass index was calculated as body weight in kilograms divided by the square of height in meters. Age, sex, treatment with blood pressure- or lipid-lowering medications or diabetes medications, and smoking status (currently smoking vs. not smoking) were assessed via questionnaire. Current smoking was defined as smoking at least one cigarette per day within the year preceding the examination. Heart rate was assessed during tonometry and was measured in beats per minute.
Menopause Assessment.
Participants were placed into subcategories based on sex and menopause age. Women were divided into four groups based on age at the last menstrual cycle: persistently premenopausal; early menopause (<25th percentile, age <48 years old); average menopause (25th to 75th percentile, age 48 to 53); or late menopause (>75th percentile, age >53). Premenopausal women who were >53 years of age at any visit were assigned to the late menopause group. Menopause status, etiology of menopause, and onset age were self-reported via questionnaires administered during the study visit. Menopause-related exclusions include unknown age of menopause and menopause induced by surgery or medication (see Supplemental Table S1).
Hemodynamic assessment.
The hemodynamic protocol has been described.3 In brief, auscultatory blood pressure was assessed in supine participants after 5 minutes of rest, followed by tonometry of the brachial, radial, femoral, and carotid arteries. Next, limited echocardiography was used to assess left ventricular outflow tract area and flow velocity, which were used to compute the time-resolved waveform of volumetric flow rate into the aorta. This volume flow waveform was combined with a calibrated carotid pressure waveform to compute Zc and perform wave separation analysis. Carotid and femoral artery tonometry, along with differences in body surface transit distance measurements from the suprasternal notch to the carotid and femoral sites, were used to calculate carotid-femoral pulse wave velocity (CFPWV). Brachial waveforms were calibrated to cuff systolic and diastolic pressure; all other waveforms were calibrated by using integrated mean and diastolic brachial pressure.
Calculation of effective internal radius and Eh of the aorta.
The physiologically effective internal radius, r, of the aorta (AoR) was calculated from the water-hammer equation, Zc=coρ/πr2, where Zc is the characteristic impedance of the aorta, co is pulse-wave velocity in cm/s, which we set equal to CFPWV, and ρ is the density of blood, which was assumed constant at 1.06 g/cm3.6 Rearranging gives: r = sqrt(coρ/πZc), where sqrt() is square root. Structural stiffness (Eh) of the aortic wall (AoEh) was then estimated by using the Moens-Korteweg equation: co2 = Eh/2rρ. Rearranging gives: Eh = 2rρco2. Good agreement between the pulse-wave velocity calculated from this equation and the true phase velocity has been demonstrated, suggesting that this is a valid method for estimating aortic radius in vivo.6
Statistical methods.
Sample clinical characteristics were summarized. Variables with skewed distributions were natural log transformed. CFPWV was transformed using the negative inverse, niCFPWV = −1000/CFPWV. Cross-sectional relations of hemodynamic measures with menopause status across three visits, adjusted for concurrent risk factors at each visit, were examined using repeated-measures linear mixed-effects models as previously described.3 Key stiffness measures for primary analyses were PP, Zc, and niCFPWV. Key derived measures were AoR and AoEh. We fitted repeated-measures linear mixed-effects models using the restricted maximum likelihood method to accommodate data that could be missing at random. We fitted models with an unstructured covariance matrix, which allows variances and correlations of the response variables to differ across visits. The repeated-measures model examined the relations between the dependent variables and concurrent age and risk factor levels (fixed effects) across the 3 repeated measurements (random effect). All evaluations at each examination cycle with a valid assessment of CFPWV, Zc, and covariates were included in these analyses, regardless of whether a participant contributed 1, 2, or 3 observations. Required covariates in base risk factor models included age (centered at 50 years), centered age squared, menopause status, heart rate, height, body-mass index, fasting blood glucose, total/HDL cholesterol ratio, triglyceride levels, status of antihypertensive, lipid-lowering or diabetes medication usage, and smoking, diabetes, or prior CVD.3 All models were adjusted for cohort.
Results
Sample characteristics
Characteristics of our sample by sex and menopause status in the 6211 participants who attended visit 1 are presented in Table 1. Characteristics of all participants at each visit, including the 549 participants who did not attend visit 1, are presented in Supplemental Tables S2 and S3. Additionally, 11 women who were persistently premenopausal but were >53 years of age at any visit were recategorized into the late menopause age subgroup since they are by definition members of that group.
Table 1.
Patient Characteristics at the Baseline Visit (N=6211)
| Variable | Men N = 3248 (52%) |
Pre-Menopausal N = 848 (13%) |
Early N = 461 (7%) |
Average N = 1171 (19%) |
Late N = 483 (8%) |
|---|---|---|---|---|---|
| Age, years | 50.5±15.5 | 32.8±7.1 | 52.6±16.1 | 53.5±13.0 | 56.8 ± 10.8 |
| Age of menopause, years | - | - | 43.5±3.7 | 50.6±1.6 | 55.5±1.7 |
| Cohort, n (%) | |||||
| Offspring | 1161 (36) | 12 (1) | 198 (43) | 463 (40) | 224 (46) |
| Omni-1 | 94 (3) | 3 (0.4) | 23 (5) | 46 (4) | 27 (6) |
| Third Generation | 1798 (55) | 748 (88) | 219 (48) | 589 (50) | 200 (41) |
| Omni-2 | 159 (5) | 84 (10) | 19 (4) | 59 (5) | 19 (4) |
| New Offspring Spouse | 36 (1) | 1 (0.1) | 2 (0.4) | 14 (1) | 13 (3) |
| MAP (mmHg) | 94.5 ± 11.1 | 84.4 ± 9.5 | 93.3 ± 12.6 | 93.2 ± 12.6 | 95.8 ± 12.7 |
| Body mass index, kg/m2 | 28 [25, 30] | 24 [21, 27] | 26 [22, 30] | 25 [22, 29] | 26 [23, 30] |
| Fasting blood glucose, mg/dL | 99 [93, 107] | 88 [83, 93] | 94 [89, 102] | 93 [88, 100] | 95 [89, 101] |
| Total to HDL cholesterol ratio | 3.9 [3.1, 4.9] | 2.9 [2.5, 3.5] | 3.1 [2.6, 3.9] | 3.0 [2.5, 3.7] | 3.1 [2.6, 3.8] |
| Triglycerides, mg/dL | 105 [73, 154] | 78 [58, 109] | 94 [64, 136] | 85 [64, 120] | 85 [64, 127] |
| Height, m | 1.76 ± 0.07 | 1.64 ± 0.06 | 1.61 ± 0.07 | 1.62 ± 0.06 | 1.63 ± 0.07 |
| HR during tonometry, BPM | 60 ± 10 | 63.1 ± 9.5 | 64.2 ± 9.6 | 61.9 ± 9.3 | 62.3 ± 9.7 |
| Prevalent CVD, n (%) | 164 (5) | 0 | 15 (3) | 25 (2) | 13 (3) |
| Lipid-lowering treatment, n (%) | 836 (26) | 11 (1) | 81 (18) | 205 (18) | 102 (21) |
| Hypertension treatment, n (%) | 873 (27) | 18 (2) | 126 (27) | 257 (22) | 128 (27) |
| Presence of diabetes, n (%) | 310 (10) | 11 (1) | 27 (6) | 63 (5) | 20 (4) |
| Current smoking, n (%) | 407 (13) | 123 (15) | 80 (17) | 101 (9) | 32 (7) |
Values are mean ± SD, n (%) or 50th [25th,75th] percentiles. The baseline visits are exam 8 for Offspring, exam 3 for Omni-1, and exam 1 for the Third generation, Omni-2 and New Offspring Spouse cohorts. MAP, mean arterial pressure; HDL, high density lipoprotein; HR, heart rate; BPM, beats per minute; CVD, cardiovascular disease.
The mean age at menopause was 44, 51, and 56 years in the early, average, and late menopause groups, respectively (Table 1). Persistently premenopausal women were 17 to 24 years younger than women in the postmenopausal groups and men. They also had lower MAP, BMI, triglycerides, and prevalence of diabetes and CVD, and were less likely to be using lipid or blood pressure-lowering medications. Men were taller and had higher BMI than women.
Age and sex relations of arterial stiffness measures
Age relations for key hemodynamic measures, adjusted for concurrent vascular risk factors, are presented in Table 2, and risk factor-adjusted age trends are presented in the Figure. This table presents results of repeated measures mixed effect models that estimate effects for centered age, age2, menopause group, and the interaction between age and menopause group. Conceptually, each of the latter interaction terms variably tilts the parabolic age relation for a given menopause group (as compared to men) in order to fit the age trend in each group and thereby determine the estimated nadir of the age relation in that group. Supplementary Table S4 and Supplemental Figure S2A–D provide minimally adjusted versions of these models. PP, FWA, and Zc had strong quadratic age relations, with risk factor-adjusted values that fell to a nadir at a variable age in midlife (Figure A–C). The locations of these nadirs were defined by the menopause group-by-age interaction terms, which all differed from the value for men (referent) but were similar across the various menopause age groups (Table 2).
Table 2.
Age relations of key hemodynamic measures.
| Pulse Pressure (mm Hg) |
Forward Wave Amplitude (mm Hg) |
Characteristic Impedance (dyne x sec/cm5) |
Carotid-Femoral PWV* (−ms/m) |
|||||
|---|---|---|---|---|---|---|---|---|
|
|
||||||||
| Variable | Est. B ± SE | p-value | Est. B ± SE | p-value | Est. B ± SE | p-value | Est. B ± SE | p-value |
| Intercept | 26.84 ± 5.24 | <0.0001 | −6.43 ± 5.02 | 0.20 | 387.9 ± 28.7 | <0.0001 | −281.83 ± 8.24 | <0.0001 |
| Age† | 0.91 ± 0.17 | <0.0001 | 0.00 ± 0.16 | 0.99 | 2.3 ± 0.9 | 0.011 | 12.91 ± 0.27 | <0.0001 |
| Age-squared | 1.55 ± 0.05 | <0.0001 | 1.61 ± 0.05 | <0.0001 | 7.6 ± 0.3 | <0.0001 | 0.66 ± 0.07 | <0.0001 |
| Menopause groups‡ | ||||||||
| Premenopausal | −0.33 ± 0.68 | 0.63 | 1.39 ± 0.64 | 0.032 | 20.9 ± 3.8 | <0.0001 | −4.75 ± 1.05 | <0.0001 |
| Early | −1.66 ± 0.57 | 0.0038 | −1.19 ± 0.55 | 0.030 | 7.6 ± 3.1 | 0.015 | −8.98 ± 0.96 | <0.0001 |
| Average | −1.86 ± 0.45 | <0.0001 | −0.71 ± 0.43 | 0.099 | 11.4 ± 2.4 | <0.0001 | −8.81 ± 0.75 | <0.0001 |
| Late | −2.09 ± 0.65 | 0.0012 | −0.78 ± 0.62 | 0.21 | 10.1 ± 3.4 | 0.0034 | −10.22 ± 1.06 | <0.0001 |
| Menopause group x age interactions‡ | ||||||||
| Premenopausal | 2.39 ± 0.39 | <0.0001 | 2.48 ± 0.37 | <0.0001 | 12.1 ± 2.1 | <0.0001 | 3.22 ± 0.57 | <0.0001 |
| Early | 2.92 ± 0.29 | <0.0001 | 1.69 ± 0.28 | <0.0001 | 8.6 ± 1.6 | <0.0001 | 1.83 ± 0.47 | <0.0001 |
| Average | 3.11 ± 0.23 | <0.0001 | 1.81 ± 0.22 | <0.0001 | 13.6 ± 1.2 | <0.0001 | 1.62 ± 0.35 | <0.0001 |
| Late | 3.18 ± 0.37 | <0.0001 | 1.80 ± 0.36 | <0.0001 | 13.5 ± 2.0 | <0.0001 | 1.92 ± 0.57 | 0.0008 |
Negative inverse transformed.
Centered at 50 years and rescaled to decades. Linear mixed models, adjusted for height, body-mass index, mean arterial pressure, heart rate, fasting blood glucose, total/high density lipoprotein cholesterol ratio, triglyceride, antihypertensive treatment, lipid-lowering medication usage, current smoking, and prevalent diabetes or cardiovascular disease. Carotid-femoral PWV was negative inverse transformed. Body-mass index, fasting blood glucose, total/high density lipoprotein cholesterol ratio, and triglyceride levels were natural log transformed.
Men are the referent group.
Figure.
Age relations for key hemodynamic measures. Plots represent the age and sex relations of (A) pulse pressure, (B) forward wave amplitude, (C) characteristic impedance, (D) negative inverse carotid-femoral pulse wave velocity, (E) proximal aortic elastance-wall thickness product, and (F) effective aortic radius in models that adjust for covariates listed in the footnote of Table 2. For the parabolic models (A-C), vertical lines represent calculated nadir values for each group predicted by the model fit. The age values corresponding to the nadirs are presented in Table 3. PP, pulse pressure; FWA, forward wave amplitude; Zc, characteristic impedance; niCFPWV, negative inverse carotid-femoral pulse wave velocity; Eh, elastance-wall thickness product.
After the midlife nadir, pulsatility measures increased with advancing age, especially in women (Figure A–C). The PP nadirs for women occurred between 37–39 years of age, across the menopause onset groups, whereas a nadir in men occurred at 47 years (Table 3). The narrow age range at the PP nadir in menopause groups led to highly variable relations between age at menopause and age at the PP nadir, which was approximately 1-2 decades before the onset of menopause (Table 3). Like PP, FWA, and Zc exhibited strong quadratic relations with age (Table 2, Figure B–C). The nadir of FWA occurred around 40-45 years of age in women compared to 50 years of age in men (Table 3, Figure B). The nadir of Zc occurred at 40–43 years of age in women, compared with 48 years of age in men (Table 3, Figure C). In contrast to pulsatility measures, CFPWV showed only modest but significant quadratic age relations (Table 2). However, the increase in CFPWV with age was steeper in each menopause group compared with men (Table 2, Figure D). In a sensitivity analysis, the age nadirs presented in Table 3 were unchanged when we did not reassign persistently premenopausal women who were >53 years of age to the late postmenopausal group. Additional sensitivity analyses examined the effects of including an interaction between menopause group and quadratic age. The resulting model had a higher Bayesian Information criterion (BIC), indicating that the more complex model was not better. In addition, we examined alternative forms of the relation between pulsatility measures and age by fitting a restricted cubic spline model for each dependent variable with 5 knots at the following percentiles of age: 5, 27.5, 50, 72.5, and 95. The resulting model for pulse pressure provided a similar pattern of estimated nadirs for women across the 4 menopause groups (41-42 years) despite a broad range of average age at onset of menopause in the 3 groups (44-56 years). The pulse pressure nadir also occurred later in men (46 years) in this alternative model.
Table 3.
Age at calculated nadir of adjusted vascular stiffness measures in groups defined by sex and menopause onset.
| Men | Pre-Menopausal | Early Menopause | Average Menopause | Late Menopause | |
|---|---|---|---|---|---|
| Pulse pressure | 47 | 39 | 38 | 37 | 37 |
| Forward wave amplitude | 50 | 42 | 45 | 44 | 44 |
| Characteristic impedance | 48 | 41 | 43 | 40 | 40 |
Nadir calculations based on linear mixed effects models presented in Table 2. All values in years.
Age and sex relations of derived values for AoEh and AoR adjusted for vascular risk factors are presented in Table 4 and Figure E–F. Supplementary Table S5 and Supplemental Figure S2E–F provide minimally adjusted versions of the models. On average, men had higher values for AoEh and AoR across the lifespan, even after adjustment for height and BMI (Figure E–F). The increase in AoEh with age was similar across men and all menopause groups, except the persistently premenopausal group, which had a slightly steeper age slope (Table 4). Aortic radius increased substantially with advancing age in men and women, starting from a similar body size-adjusted value in young adults, with men having a steeper increase with age through midlife (Table 4, Figure F). Across all groups of women, AoR converged on an apparent ceiling at an earlier age and at a lower value than in men (Figure F). The attenuated increase in AoR before midlife in women is consistent with the attenuated reduction in pressure pulsatility measures from young adulthood through midlife (Figure A–C). Similarly, the attenuated increase with age and lower ceiling for AoR in women after midlife is consistent with the steeper rise in pulsatility measures (Figure A–C), which are highly sensitive to AoR, and the modestly greater increase in CFPWV (Figure D), which is much less sensitive to AoR, despite a comparable rise in Eh (Figure E) with age in women as compared to men.
Table 4.
Age relations of aortic structural stiffness and effective lumen radius.
| Aortic Eh (ln(dyne/cm)) |
Aortic Radius (mm) |
|||
|---|---|---|---|---|
|
|
||||
| Variable | Est. B±SE | p-value | Est. B±SE | p-value |
| Intercept | 10.706 ± 0.190 | <0.0001 | −0.278 ± 0.837 | 0.74 |
| Age† | 0.275 ± 0.006 | <0.0001 | 0.624 ± 0.027 | <0.0001 |
| Age-squared | 0.022 ± 0.002 | <0.0001 | −0.071 ± 0.008 | <0.0001 |
| Menopause groups‡ | ||||
| Premenopausal | −0.089 ± 0.025 | 0.0003 | −0.666 ± 0.108 | <0.0001 |
| Early | −0.184 ± 0.022 | <0.0001 | −0.635 ± 0.092 | <0.0001 |
| Average | −0.186 ± 0.017 | <0.0001 | −0.719 ± 0.072 | <0.0001 |
| Late | −0.227 ± 0.024 | <0.0001 | −0.852 ± 0.103 | <0.0001 |
| Menopause group x age interactions‡ | ||||
| Premenopausal | 0.037 ± 0.013 | 0.005 | −0.128 ± 0.061 | 0.036 |
| Early | −0.003 ± 0.011 | 0.78 | −0.122 ± 0.047 | 0.009 |
| Average | −0.014 ± 0.008 | 0.087 | −0.221 ± 0.036 | <0.0001 |
| Late | −0.015 ± 0.013 | 0.26 | −0.221 ± 0.059 | 0.0002 |
Centered at 50 years and rescaled to decades. Linear mixed models, adjusted for height, body-mass index, mean arterial pressure, heart rate, fasting blood glucose, total/high density lipoprotein cholesterol ratio, triglyceride, antihypertensive treatment, lipid-lowering medication usage, current smoking, and prevalent diabetes or cardiovascular disease. Body-mass index, fasting blood glucose, total/high density lipoprotein cholesterol ratio, and triglyceride levels were natural log transformed. Eh, Young’s modulus x wall thickness product.
Men are the referent group.
Discussion
We examined age relations of multiple related but distinct measures of aortic stiffness and pressure pulsatility, including PP, FWA, Zc, and CFPWV, in men and women in the community-based Framingham Heart Study. Using these measures, we computed 2 derived variables, AoEh and AoR, that characterize the relations among wall stiffness, lumen radius, and aortic hemodynamic properties. We observed that a previously described midlife PP nadir occurred approximately a decade earlier in women compared to men. Importantly, when women were categorized by age at menopause onset (last menstrual cycle), there were only slight differences in age at the PP nadir between groups. Age relations of FWA and Zc resembled those of PP. In parallel, the derived measure of AoEh showed similar monotonically increasing age relations in men and women, with higher values in men at all ages. A derived measure of average aortic radius (AoR), adjusted for body size and risk factors, was similar in young men and women but became progressively larger in men than in women across the lifespan. AoR increased rapidly through midlife and continued to increase into the eighth decade of life in men. In contrast, AoR increased more slowly with age and plateaued earlier in women across all menopause groups, leading to progressive divergence in AoR values between men and women. Taken together, these findings indicate that the midlife transition from falling to rising pressure pulsatility and related measures begins earlier in life in women than in men, and well before the traditional onset of menopause (last menstrual cycle), underscoring the contribution of factors unrelated to the end of ovulation in shaping this age-related component of vascular health.
Age relations of pressure pulsatility measures differ from those of aortic wall stiffness. Differences in aortic diameter are known to mediate the dissociation between measures of aortic wall stiffness (CFPWV) and pressure pulsatility (PP, FWA, Zc).3,5,7,14 We estimated average lumen radius (AoR) and structural wall stiffness (Eh) based on the observed hemodynamic measures. Prior studies at FHS and in Iceland have demonstrated remarkably similar sex differences in the age-related changes in directly measured aortic diameter. For example, Lam et al. examined repeated direct assessments of aortic root diameter in FHS participants in midlife and reported a lower average diameter and a lower age-related longitudinal increase in aortic diameter in women than in men.5,15 They also noted that age-related changes in PP were negatively associated with changes in aortic root diameter, consistent with a known strong inverse relation between aortic lumen diameter and characteristic impedance and pressure pulsatility.3,7,15–18
Age relations of pulsatility and stiffness measures differed modestly across menopause groups in our sample, despite marked differences in menopausal age, suggesting that the mechanisms underlying sex differences in hemodynamic changes with age are distinct from menopausal age. The presence of a PP nadir years before the end of menstrual cycles could be consistent with perimenopause, with related alterations in levels of estrogen and other hormones.19 However, the convergent timing of the PP nadir in the premenopausal and 3 postmenopausal groups suggests that additional sex differences, unrelated to age of final menstrual period, are contributing to the earlier hemodynamic transition in women as compared to men.
Prior clinical studies that examined relations of menopause and vascular markers are limited and report mixed findings. Several cross-sectional studies have demonstrated that early natural or surgical menopause is associated with increased risk for atherosclerotic CVD in women.20–22 One recent study compared normal-onset and late-onset post-menopausal women and examined total aortic stiffness by measuring pulse wave velocity and blood levels of lipid-related metabolites known to induce endothelial dysfunction and reactive oxygen species. They reported that women with late-onset menopause have lower total aortic stiffness and distinct lipid-related metabolite profiles, suggesting that the timing of menopause is an important predictor of age-related vascular aging. 23
Longitudinal studies evaluating vascular trajectories across a woman’s lifespan are also lacking. Most existing longitudinal studies focus on the period immediately surrounding menopause and perimenopause; however, this narrow time frame may be insufficient to capture vascular changes that emerge earlier in life, particularly given evidence that the menopausal transition may begin as early as the mid-thirties to early forties.9 Therefore, current cross-sectional and longitudinal studies are insufficient to fully characterize the changes that occur during a woman’s lifetime.
A recent analysis examined relations between menopause onset and vascular stiffness over a 14-year interval in the Longitudinal Assessment of Women (LAW) study in Australia. The LAW investigators reported that the age slope of CFPWV was not significantly different between premenopause/menopause transition and early menopause groups but was marginally steeper in the late menopause group, concluding that the late menopause group experiences an acceleration in the age slope of CFPWV >8 years after menopause onset.12 We observed progressive convergence of CFPWV values with age in women and men, though men maintained higher values throughout life. The convergence is due to steeper age slopes in women than in men, but unlike the LAW study, these age slopes were similar across the postmenopausal groups in our study (Table 2, Figure 1D). In addition, the LAW investigators examined untransformed CFPWV, which is known to increase nonlinearly with age and to exhibit considerable heteroskedasticity after midlife, perhaps contributing to the apparently steeper age slope in the late menopause group.
While data are mixed regarding menopausal age and the age relations of aortic stiffness measures in women, they do support a role for estrogen in modulating endothelial function and atherosclerosis progression, which may depend on the years since menopause at the time of initiation of treatment.24 Estrogen has also been shown to modulate endothelial nitric oxide production, endothelial signaling, and oxidative stress, which supports in-vitro and ex-vivo influence of estrogen on vascular health.8,24,25 This is further supported by randomized clinical studies showing that exogenous estrogen administration in postmenopausal women can improve aortic stiffness measures.24,26–28
Taken together, existing evidence supports a protective influence of estrogen on the vascular endothelium and overall cardiovascular health. While cross-sectional and short-term longitudinal studies suggest that menopausal fluctuations in estrogen levels may contribute to increases in vascular stiffness, our findings indicate that key changes in vascular stiffness emerge much earlier in life. The earlier transition from falling to rising pressure pulsatility in women as compared to men and the consistent nadir in pulsatility measures well before menopause, regardless of timing of menopause onset, suggest that sex differences in the age relations of vascular measures are distinct from hormonal changes occurring at or after the onset of menopause. However, effects of fluctuating estrogen levels, with associated vasomotor symptoms, during the perimenopausal period cannot be excluded. While further studies that directly measure sex hormones over time are necessary to confirm this hypothesis, the onset of menopause and associated changes in estrogen levels may modulate vascular function or exert short-term effects but may not be solely responsible for the accelerated increase in pressure pulsatility in women. Rather, additional underlying mechanisms may contribute well before the midlife menopausal transition.
There are several possible non-hormonal contributors to sex differences in PP. First, prior studies that measured proximal aortic diameter directly have suggested that late-life sex differences in pulsatility measures are related to differences in measured diameter of the proximal aorta.5,7 Derived AoR, which was computed from Zc and CFPWV in the present analysis and adjusted for body size, plateaued earlier and converged on a substantially lower level in women as compared to men. The foregoing findings using either measured or derived aortic diameter are consistent with intrinsic sex differences in aortic structure and remodeling with advancing age. Considering similar age relations of measures of aortic wall stiffness between the sexes, differences in aorta size may offer a mechanical explanation for the differing trends in PP after midlife.
Another possible mechanism for sex differences in the age relations of hemodynamic measures may be differences in levels of or responsiveness to non-estrogen hormones. N-terminal pro-B-type natriuretic peptide (NT-proBNP), which promotes diuresis, vasodilation, and inhibition of the renin-angiotensin-aldosterone system, is correlated with sex hormone levels. Levels of NT-proBNP fall substantially in women after menopause, resulting in older women having a relative natriuretic peptide handicap.29–32 Finally, other sex-related factors, such as parity, may affect vascular stiffness, though further studies are required to examine these relations.
Our study has several limitations. First, we did not measure the proximal aortic diameter directly but rather assessed the effective aortic diameter by examining relations between Zc and CFPWV. Prior analyses using this approach yielded identical insights to those obtained in analyses of directly measured proximal aortic diameter.3,16 Second, our study did not measure vascular stiffness at pre-determined times before and after menopause but instead used a repeated measures approach, which limits our ability to precisely describe differences in hemodynamic measures immediately before and after the menopause transition in individual participants. However, the goal of the analysis was to observe differences in trajectories of hemodynamic measures according to early, average, and late onset of menopause. Due to the longitudinal nature and large sample size, our study is better equipped to observe these age trends over time. Third, menopause-timing was based on a self-reported age rather than direct measurements of sex hormones, which would be needed to objectively establish estrogen exposure levels prior to onset of menopause. Future studies that relate hormonal changes over time to clinical features of menopause and changes in vascular measures are required. Fourth, our sample was largely composed of individuals of White European descent. External validation of our results in other racial and ethnic groups will be required. Fifth, our study excluded 22% of women due to menopause-related exclusions, which may have affected our results. Sixth, the persistently premenopausal group is by definition younger and healthier than the other subgroups. That being said, the persistently premenopausal group serves to balance the global trends in pulsatility measures over time across the full age range in women as compared to men. Lastly, our study was observational, and we cannot exclude residual confounding or establish causal relations
Perspectives
We have demonstrated an important dissociation in the midlife onset of increasing PP and onset of menopause, with the transition to increasing PP occurring an average of 6-19 years before menopause. The pattern of marked differences in changes in measures of pressure pulsatility, compared with measures of aortic wall stiffness, suggests aortic diameter remodeling as a relating factor and signals a limited capacity for aortic diameter remodeling in later life in women. Strong relations of vascular stiffness with CVD risk in women argue for aggressive preventive management in younger women, including mindful assessment of PP, and further exploration of factors contributing to marked sex differences in PP-age relations and lifestyle modifications or pharmacologic interventions that can prevent or attenuate the unfavorable increase in PP after midlife.
Supplementary Material
Novelty and Relevance.
What Is New?
Women have lower pulse pressure (PP) than men prior to midlife, when PP is falling with age, but higher PP after midlife. Sex differences in PP have been attributed to loss of estrogen in women after menopause. However, we found that PP and related vascular measures transitioned from falling to increasing with age a decade earlier in women than men and up to 2 decades before the onset of menopause.
What is Relevant?
Sex differences in aortic diameter remodeling, which is blunted in women, contribute to the marked increase in PP in women after midlife.
Clinical/Pathophysiological Implications?
Our observations highlight the need for additional work to elucidate factors that contribute to the increase in PP and cardiovascular disease risk in women after midlife.
Acknowledgments
We thank the FHS participants, without whom this research would not be possible.
Funding Sources
This study was supported by NHLBI contract 75N92019D00031 (R.S.V) and by DK080739 (R.S.V), HL107385, HL126136, HL93328, HL142983, HL143227, HL131532 and AG079390 (R.S.V., G.F.M.), and HL70100, HL092577, 2U54HL120163, AG066010 and K12AR085635 (E.J.B.). R.S.V. was supported in part by the Evans Medical Foundation and the Jay and Louis Coffman Endowment from the Department of Medicine, Boston University Chobanian and Avedisian School of Medicine. N.M.H. was funded by U54HL120163, AHA 20SRFRN35120118, R01HL115391, and R01HL168889. L.L.C. was funded by NHLBI (K01HL161494).
Disclosures
G.F.M. is the owner of Cardiovascular Engineering, Inc., a company that designs and manufactures devices that measure vascular stiffness. The company uses these devices in clinical trials that evaluate the effects of diseases and interventions on vascular stiffness. G.F.M. also serves as a consultant to and receives grants and honoraria from Novartis, Merck, Bayer, Servier, Philips, and deCODE genetics. G.F.M. is an inventor on a pending patent application that discloses methods for predicting various measures of biological age using pressure waveforms. G.F.M. is a co-inventor on a pending patent application that discloses a method for estimating carotid-femoral pulse wave velocity and vascular age by using a convolutional neural network. The remaining authors report no conflicts.
Non-standard Abbreviations:
- PP
Pulse pressure
- FWA
Forward wave amplitude
- Zc
Aortic characteristic impedance
- CFPWV
Carotid-femoral pulse wave velocity
- niCFPWV
Negative inverse carotid-femoral pulse wave velocity
- AoR
Aortic radius
- AoEh
Aortic elastance x wall thickness product
References:
- 1.Madhavan S, Ooi WL, Cohen H, Alderman MH. Relation of pulse pressure and blood pressure reduction to the incidence of myocardial infarction. Hypertension. 1994;23:395–401. doi: 10.1161/01.HYP.23.3.395 [DOI] [PubMed] [Google Scholar]
- 2.Mitchell GF, Wang N, Palmisano JN, Larson MG, Hamburg NM, Vita JA, Levy D, Benjamin EJ, Vasan RS. Hemodynamic correlates of blood pressure across the adult age spectrum: noninvasive evaluation in the Framingham Heart Study. Circulation. 2010;122:1379–1386. doi: 10.1161/CIRCULATIONAHA.109.914507 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Mitchell GF, Lacourcière Y, Ouellet JP, Izzo JL Jr, Neutel J, Kerwin LJ, Block AJ, Pfeffer MA. Determinants of elevated pulse pressure in middle-aged and older subjects with uncomplicated systolic hypertension: the role of proximal aortic diameter and the aortic pressure-flow relationship. Circulation. 2003;108:1592–1598. doi: 10.1161/01.CIR.0000093435.04334.1F [DOI] [PubMed] [Google Scholar]
- 4.Mitchell GF, Hwang SJ, Vasan RS, Larson MG, Pencina MJ, Hamburg NM, Vita JA, Levy D, Benjamin EJ. Arterial stiffness and cardiovascular events. Circulation. 2010;121:505–511. doi: 10.1161/CIRCULATIONAHA.109.886655 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Lam CSP, Xanthakis V, Sullivan LM, Lieb W, Aragam J, Redfield MM, Mitchell GF, Benjamin EJ, Vasan RS. Aortic root remodeling over the adult life course: longitudinal data from the Framingham Heart Study. Circulation. 2010;122:884–890. doi: 10.1161/CIRCULATIONAHA.110.937839 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Mitchell GF, Rong J, Larson MG, Hamburg NM, Vita JA, Levy D, Benjamin EJ, Vasan RS. Longitudinal Hemodynamic Correlates of and Sex Differences in the Evolution of Blood Pressure Across the Adult Lifespan: The Framingham Heart Study. J Am Heart Assoc. 2023;12. doi: 10.1161/JAHA.122.027329 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Mitchell GF, Gudnason V, Launer LJ, Aspelund T, Harris TB. Hemodynamics of increased pulse pressure in older women in the community-based Age, Gene/Environment Susceptibility-Reykjavik Study. Hypertension. 2008;51:1123–1128. doi: 10.1161/HYPERTENSIONAHA.107.108175 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Chambliss KL, Shaul PW. Estrogen Modulation of Endothelial Nitric Oxide Synthase. Endocr Rev. 2002;23:665–686. doi: 10.1210/er.2001-0045 [DOI] [PubMed] [Google Scholar]
- 9.Samargandy S, Matthews KA, Brooks MM, Barinas-Mitchell E, Magnani JW, Janssen I, El Khoudary SR. Arterial Stiffness Accelerates Within 1 Year of the Final Menstrual Period. Arterioscler Thromb Vasc Biol. 2020;40:1001–1008. doi: 10.1161/ATVBAHA.119.313622 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Hildreth KL, Ozemek C, Kohrt WM, Blatchford PJ, Moreau KL. Vascular dysfunction across the stages of the menopausal transition is associated with menopausal symptoms and quality of life. Menopause. 2018;25:1011–1019. doi: 10.1097/GME.0000000000001112 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Moreau KL. Intersection between gonadal function and vascular aging in women. J Appl Physiol. 2018;125:1881–1887. doi: 10.1152/japplphysiol.00117.2018 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.O’Neill SM, Travers CM, Otahal P, Khoo SK, Sharman JE. Menopause and accelerated aortic stiffness. Maturitas. 2024;180. doi: 10.1016/j.maturitas.2023.107900 [DOI] [PubMed] [Google Scholar]
- 13.Takahashi K, Miura S, Mori-Abe A, Kawagoe J, Takata K, Ohmichi M, Kurachi H. Impact of menopause on the augmentation of arterial stiffness with aging. Gynecol Obstet Invest. 2005;60:162–166. doi: 10.1159/000086570 [DOI] [PubMed] [Google Scholar]
- 14.Mitchell GF, Solomon SD, Shah AM, Claggett B, Rouleau JL, Sweitzer NK, O’Meara E, Desai AS, Anand IS, Packer M, et al. Hemodynamic Effects of Sacubitril-Valsartan Versus Enalapril in Patients With Heart Failure in the EVALUATE-HF Study: Effect Modification by Left Ventricular Ejection Fraction and Sex. Circ Heart Fail. 2021;14:e007891. doi: 10.1161/CIRCHEARTFAILURE.120.007891 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Mitchell GF, Pfeffer MA, Finn PV, Pfeffer JM. Equipotent antihypertensive agents variously affect pulsatile hemodynamics and regression of cardiac hypertrophy in spontaneously hypertensive rats. Circulation. 1996;94:2923–2929. doi: 10.1161/01.cir.94.11.2923 [DOI] [PubMed] [Google Scholar]
- 16.Mitchell GF, Conlin PR, Dunlap ME, Lacourcière Y, Arnold JMO, Ogilvie RI, Neutel J, Izzo JL Jr, Pfeffer MA. Aortic diameter, wall stiffness, and wave reflection in systolic hypertension. Hypertension. 2008;5:105–111. doi: 10.1161/HYPERTENSIONAHA.107.099721 [DOI] [PubMed] [Google Scholar]
- 17.Vasan RS, Larson MG, Levy D. Determinants of echocardiographic aortic root size. The Framingham Heart Study. Circulation. 1995;91:734–740. doi: 10.1161/01.cir.91.3.734 [DOI] [PubMed] [Google Scholar]
- 18.Torjesen AA, Sigurðsson S, Westenberg JJM, Gotal JD, Bell V, Aspelund T, Launer LJ, Harris TB, Gudnason V, Mitchell GF. Pulse pressure relation to aortic and left ventricular structure in the Age, Gene/Environment Susceptibility (AGES)-Reykjavik Study. Hypertension. 2014;64:756–761. doi: 10.1161/HYPERTENSIONAHA.114.03870 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Laakkonen EK, Karppinen JE, Lehti S, Lee E, Hämäläinen M, Kovanen V, Aukee P, Sipilä S, Kaprio J, Ollikainen M, et al. Associations of Sex Hormones and Hormonal Status With Arterial Stiffness in a Female Sample From Reproductive Years to Menopause. Front Endocrinol (Lausanne). 2021;12:765916. doi: 10.3389/fendo.2021.765916 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Zhu D, Chung HF, Dobson AJ, Pandeya N, Giles GG, Bruinsma F, Brunner EJ, Kuh D, Hardy R, Avis NE, et al. Age at natural menopause and risk of incident cardiovascular disease: a pooled analysis of individual patient data. Lancet Public Health. 2019;4:e553–e564. doi: 10.1016/S2468-2667(19)30155-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Muka T, Oliver-Williams C, Kunutsor S, Laven JSE, Fauser BCJM, Chowdhury R, Kavousi M, Franco OH. Association of Age at Onset of Menopause and Time Since Onset of Menopause With Cardiovascular Outcomes, Intermediate Vascular Traits, and All-Cause Mortality: A Systematic Review and Meta-analysis. JAMA Cardiol. 2016;1:767–776. doi: 10.1001/jamacardio.2016.2415 [DOI] [PubMed] [Google Scholar]
- 22.Honigberg MC, Zekavat SM, Aragam K, Klarin D, Bhatt DL, Scott NS, Peloso GM, Natarajan P. Association of Premature Natural and Surgical Menopause With Incident Cardiovascular Disease. JAMA. 2019;322:2411–2421. doi: 10.1001/jama.2019.19191 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Darvish S, Murray KO, Mahoney SA, et al. Late-Onset Menopause and Attenuated Aortic Stiffness. Hypertension. Published online January 12, 2026. doi: 10.1161/HYPERTENSIONAHA.125.26346 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Hodis HN, Mack WJ, Henderson VW, Shoupe D, Budoff MJ, Hwang-Levine J, Li Y, Feng M, Dustin L, Kono N, et al. Vascular Effects of Early versus Late Postmenopausal Treatment with Estradiol. N Engl J Med. 2016;374:1221–1231. doi: 10.1056/NEJMoa1505241 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.SenthilKumar G, Katunaric B, Bordas-Murphy H, Sarvaideo J, Freed JK. Estrogen and the Vascular Endothelium: The Unanswered Questions. Endocrinology. 2023;164. doi: 10.1210/endocr/bqad079 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Matsui S, Yasui T, Tani A, Kunimi K, Uemura H, Yamamoto S, Kuwahara A, Matsuzaki T, Irahara M. Effect of ultra-low-dose estradiol and dydrogesterone on arterial stiffness in postmenopausal women. Climacteric. 2014;17:191–196. doi: 10.3109/13697137.2013.856399 [DOI] [PubMed] [Google Scholar]
- 27.Miura S, Tanaka E, Mori A, Yasui T, Komatsu T, Ushiroyama T, Sakuma K, Ito M. Hormone replacement therapy improves arterial stiffness in normotensive postmenopausal women. Maturitas. 2003;45:293–298. doi: 10.1016/S0378-5122(03)00158-0 [DOI] [PubMed] [Google Scholar]
- 28.Karim R, Xu W, Kono N, Mack WJ, Hodis HN. Effect of menopausal hormone therapy on arterial wall echomorphology: Results from the Early versus Late Intervention Trial with Estradiol (ELITE). Maturitas. 2022;162:15–22. doi: 10.1016/j.maturitas.2022.02.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Lam CSP, Cheng S, Choong K, Larson MG, Murabito JM, Newton-Cheh C, Bhasin S, McCabe EL, Miller KK, Redfield MM, et al. Influence of sex and hormone status on circulating natriuretic peptides. J Am Coll Cardiol. 2011;58:618–626. doi: 10.1016/j.jacc.2011.03.042 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Redfield MM, Rodeheffer RJ, Jacobsen SJ, Mahoney DW, Bailey KR, Burnett JC Jr. Plasma brain natriuretic peptide concentration: impact of age and gender. J Am Coll Cardiol. 2002;40:976–982. doi: 10.1016/S0735-1097(02)02059-4 [DOI] [PubMed] [Google Scholar]
- 31.Glisic M, Rojas LZ, Asllanaj E, Vargas KG, Kavousi M, Ikram MA, Voortman T, Franco OH, Muka T. Sex steroids, sex hormone-binding globulin and levels of N-terminal pro-brain natriuretic peptide in postmenopausal women. Int J Cardiol. 2018;261:189–195. doi: 10.1016/j.ijcard.2018.03.008 [DOI] [PubMed] [Google Scholar]
- 32.Chang AY, Abdullah SM, Jain T, Stanek HG, Das SR, McGuire DK, Khera A, de Lemos JA. Associations among androgens, estrogens, and natriuretic peptides in young women: observations from the Dallas Heart Study. J Am Coll Cardiol. 2007;49:109–116. doi: 10.1016/j.jacc.2006.10.040 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.